Introduction
South Africa’s colonial and apartheid systems were characterized by inequality and discrimination that was systematic, extensive and pervaded almost all aspects of public and private life. This is the context that informs South Africa’s imposition of the anti-discrimination duty not only on the state, but also on private persons. Section 9(4) of the Constitution prohibits all persons from unfairly discriminating against anyone, and requires the state to enact national legislation to prevent or prohibit unfair discrimination. The Promotion of Equality and Prevention of Unfair Discrimination Act 4 of 2000 (PEPUDA) is this legislation. It affirms the broad class of anti-discrimination duty-bearers and, in section 14, puts in place a test for determining whether discrimination is unfair, which requires a court to evaluate both the impact of that discrimination on the complainant and any possible justifications put forward by the discriminator.
While both the state and private persons bear anti-discrimination duties, there are some important ways in which private persons, and the private sphere, differ from the state and public entities (construed broadly), which complicates the adjudication of disputes between private persons in anti-discrimination law. Most significantly, the basis for justifying discrimination may sometimes be larger for a private person than it is for the state. Until recently, how anti-discrimination duties play out in the law of succession, particularly for private testamentary bequests, was unclear. On the one hand, a commitment to personal autonomy and the public / private divide might suggest that differentiating between bequests that are public in character – say, charitable trusts – and those that are private is normatively justifiable. On the other, as both a fundamental right and a central value in South Africa’s constitutional order, equality has operated to infuse and constrain private relations, and so would exert force on the law of succession too.
This question recently came to a head in two Constitutional Court judgments: King v De Jager and Wilkinson v Crawford.Footnote 1 Both concerned discriminatory provisions in private testamentary instruments, and so pitted freedom of testation (and implicated rights to privacy, dignity and property) against the right to equality. In both judgments, a majority of the Court held that freedom of testation is circumscribed by the demands of equality, though the judges differed on the relevant test, the extent to which duties borne by private persons mirror those of the state, and the balance to be struck when private persons’ rights and interests compete.
This article has four substantive parts. I begin by sketching the approach to freedom of testation prior to these judgments. Next, I discuss the King and Wilkinson judgments in some detail, before mapping out the divergent approaches in the majority and minority judgments and engaging with criticism of the judgments, including that they constitute an impermissible incursion into private autonomy and property. I argue instead that the judgments’ outcomes are warranted in a constitutional regime that sets its face against private forms of discrimination. However, I contend that how the Constitutional Court extended these duties raises more questions than it resolves, not least because it risks perpetuating parallel legal paths. Finally, I suggest that a harmonized approach should be taken to the adjudicative application question, bringing together section 14 of PEPUDA with the common law public policy test. I do not approach these judgments as a specialist in the law of succession, but rather because they exemplify how anti-discrimination duties and substantive equality play out in the private sphere.
The common law background
In this part, I briefly consider freedom of testation, its content and its limits as developed in case law. Although there is no express constitutional provision protecting freedom of testation, it is generally accepted, including by the Constitutional Court, that the principle forms part of the rights to dignity, to privacy and against arbitrary deprivation of property protected in sections 10, 14 and 25 of the Constitution.Footnote 2 Freedom of testation also instantiates a (negative) autonomy interest, a right to have one’s personal decisions be free from state intrusion and to exercise choice over decisions of personal importance.Footnote 3 Freedom of testation, then, has indirect constitutional protection.Footnote 4
However, this freedom is not without limits. As du Toit notes, public policy has long played the role of circumscribing the bounds of societally acceptable bequests in the law of succession.Footnote 5 The seemingly private sphere was never entirely insulated from judicial intervention or regulation – clear in the regular invalidation by the courts of testamentary conditions aimed at the destruction of marriagesFootnote 6 – although the courts recognized conditions which prohibited marriage on the grounds of religion, nationality or race as valid.Footnote 7 Any bequest contrary to public policy would not be enforced by courts, although pre-constitutional case law demonstrates that courts were sometimes reluctant to apply that test to limit freedom of testation where charitable testamentary bequests were based on race, gender, nationality or religion.Footnote 8 Now, freedom of testation is subject to the Constitution’s demands, including the right to equality. Du Toit, pointing out that freedom of testation has never been completely unfettered in South African law, argues that the courts should adopt a “casuistic though principled approach” to balancing competing rights.Footnote 9
Prior to the Constitutional Court’s judgments in King and Wilkinson, several testamentary provisions were tested in lower courts against the Constitution’s demands.Footnote 10 In Minister of Education v Syfrets Trust Ltd NO, a 2006 judgment, the Western Cape High Court invalidated terms of a public trust that restricted bursaries to men of “European descent” who were not Jewish, on the basis that this bequest was contrary to public policy under the Constitution. The judgment held that, even if freedom of testation is understood as a proprietary interest protected by section 25, in this instance the constitutional protection of property rights only protects against the arbitrary deprivation of property. The judgment engaged in a balancing test and found that the right to equality outweighs the right to private property and privacy.Footnote 11 However, the judgment was careful to emphasize the public dimensions at play; after all, the case involved a public trust, and the terms of the testamentary instrument were to be enforced by an organ of state (a university). Commentators have noted that the judgment does not offer extensive guidance on more difficult cases involving private wills or trusts, or less patently unfair discriminatory provisions.Footnote 12
In 2012’s Board of Executors v Benjamin Godlieb Heydendrych Testamentary Trust, George King Testamentary Trust, Cyril Houghton Bursary Trust, Women’s Legal Centre (Amicus Curiae), the Western Cape High Court again found that educational trusts that restricted benefits on the grounds of race and gender were unfairly discriminatory.Footnote 13 Similarly, in Curators Ad Litem to Certain Potential Beneficiaries of Emma Smith Educational Fund v The University of KwaZulu-Natal in 2010, the Supreme Court of Appeal ordered that a provision that limited university bursaries to white women be deleted, on the basis that it contravenes public policy and the right to equality, which cannot be countenanced in a public trust.Footnote 14 Two years later, however, in BOE Trust Ltd NO (in their capacities as co-trustees of the Jean Pierre De Villiers Trust), the Supreme Court of Appeal declined to find that a testamentary provision that limited bursaries to white South Africans was contrary to public policy, on the basis that it was impossible to fulfil (as the universities refused to administer such a bursary), and so in any event the bequest defaulted to an identified charity. The Supreme Court of Appeal distinguished its earlier precedent on the basis that the testator in Emma Smith did not provide any alternatives in the event of impossibility of performance.Footnote 15 This line of reasoning has not been without criticism: Modiri castigates the judgment’s failure to directly confront the discriminatory provisions as racially conservative.Footnote 16
These earlier judgments show that courts were reluctant to enforce discriminatory provisions in certain testamentary contexts. The courts employed the common law’s public policy test to integrate constitutional equality imperatives.Footnote 17 Significantly, the judgments all involved a public dimension; in three of them, courts declared the provisions invalid and contrary to public policy for violating the right to equality.Footnote 18 In one judgment, the Supreme Court of Appeal sidestepped the question on the basis that the provision was impossible to fulfil.Footnote 19 This tracked an enduring, if not necessarily comprehensive, commitment to the public / private divide, in which freedom of testation was seen to give expression to autonomy, understood to fall in the heartland of the private sphere.Footnote 20 However, section 9(4) of the Constitution and PEPUDA’s prohibition of unfair discrimination by private persons shifts the needle: the anti-discrimination duty is not only imposed on the state, but also on private persons. It was an open question, then, whether the Constitutional Court would extend the reasoning in these earlier cases to the private sphere.
If yes, the Court would also need to consider how the anti-discrimination duty should be imposed, and the role of both common law public policy and the unfair discrimination enquiry. South Africa’s equality provisions prohibit unfair discrimination, and so substantively, much turns on the determination of fairness. Harksen v Lane, South Africa’s locus classicus on unfair discrimination, was decided in 1997, prior to the passage of PEPUDA.Footnote 21 The court set out a three-stage test:
1. whether a provision differentiates between persons on the basis of a rational connection;
2. whether the differentiation amounts to unfair discrimination, which in turn requires examining–
i. whether the differentiation amounts to discrimination; and
ii. whether if yes, it amounts to unfair discrimination, with the unfairness test focusing primarily on the impact of the discrimination on the complainant and others in their situation; and
3. if the discrimination is unfair, whether it can nevertheless be justified under the limitations analysis.
The Harksen test has been criticized for rendering the limitations analysis obsolete.Footnote 22 The remaining heart of the test turns on the fairness or unfairness of the discrimination. PEPUDA’s section 14, enacted after Harksen v Lane, was partially based on the judicial factors proposed in that judgment, as well as components of the limitations analysis.Footnote 23 Section 14 folds in both an impact assessment, focusing on the effects on the complainant (including whether the discrimination perpetuates existing inequalities), and a justification assessment, focusing on the reasons proffered by the discriminator. However, it has remained largely doctrinally untested, and King and Wilkinson provided the opportunity for the Constitutional Court to give the section meaning and harmonize it with the existing common law.
Breaking ground: Two doctrinal developments
I now turn to two significant recent Constitutional Court judgments on the scope of anti-discrimination duties in the private sphere: King and Wilkinson. Both judgments concern the law of succession: in both, the applicants relied on their rights not to be discriminated against unfairly (on the grounds of gender and birth respectively) to vindicate inheritance claims. In both, the applicants were successful, with the Court affirming that the right not to be discriminated against constrains freedom of testation.
King v De Jager
In King, the Constitutional Court found that a private testator cannot discriminate unfairly on the grounds of sex and gender. A 1902 will established a fideicommissum bequeathing farms to male descendants only, until a third generation.Footnote 24 When two third-generation heirs died without sons, this clause prevented their female descendants from inheriting, despite one heir purporting to leave his share to his five daughters. The female descendants challenged this clause, arguing that it constituted unfair discrimination based on gender and sex. Their cause of action relied on the constitutional right and PEPUDA, particularly section 8, which prohibits unfair discrimination on the ground of gender, including through “the system of preventing women from inheriting family property” (section 8(c)) and “any practice, including traditional, customary or religious practice, which impairs the dignity of women and undermines equality between women and men” (section 8(d)).
Both the High Court and the Supreme Court of Appeal dismissed the granddaughters’ application.Footnote 25 The High Court reasoned that the clause, as a once-off private disposition, did not engage any system or practice as stipulated in section 8 of PEPUDA.Footnote 26 Bozalek J also held that any infringement of the right to equality was justified, and that this right should not trump property, in part because this would generate an arbitrary result (as prior female generations had not benefited).Footnote 27 The Supreme Court of Appeal dismissed the case after a hearing without – rather astonishingly – providing any reasons at all.
On appeal, the Constitutional Court handed down three judgments, all agreeing on the outcome but differing in their reasoning. While the respondents (the grandsons and great-grandsons) initially pressed their claim, in oral argument they conceded that the provision was unfairly discriminatory. This concession was fatal: it enabled the Court’s majority judgment (per Jafta J) to bypass a crucial component of the analysis – that is, whether the discrimination could be justified as fair. On the basis of the concession, the Court declared the clause that restricted the fideicommissum to male descendants inconsistent with the Constitution and PEPUDA and therefore unenforceable.
However, the three judgments differed on the normative reach of the prohibition on private discrimination and on how to test the constitutionality of the testamentary clause. First, Jafta J (the majority judgment) held that section 8 of PEPUDA applies, with the effect that the discrimination is unfair. Importantly, held Jafta J, the respondents’ concession was determinative of the case, and so his judgment bypasses any justification analysis.Footnote 28 Second, Mhlantla J’s separate judgment found that the provision was unenforceable based on public policy, advocating for the indirect application of the right to equality via developing the common law, rather than through PEPUDA.Footnote 29 Third, Victor AJ wrote separately to find that the common law need not be developed and that instead PEPUDA’s provisions should be applied.Footnote 30 She notes that the common law should only be developed if legislation does not give effect to the right. As in its recent Pridwin judgment (on horizontal application and the constitutional duties of private schools), much of the Court’s attention was taken up by debates about the form of the application of rights to private persons – that is, how the prohibition against private discrimination should be imposed through applicable legal rules.Footnote 31 The Court was thus divided on whether the duties on the testator are imposed indirectly, through developing the common law test of public policy, through the subsidiary provisions of PEPUDA, or directly based on provisions of the Constitution. As I argue below, this division risks proliferating parallelism in South African law.
The Court also differed on the relevant test for determining how (or indeed, whether) to balance competing private interests. Jafta J’s majority judgment found the testamentary provision unfairly discriminatory and thus not protected by freedom of testation.Footnote 32 Quite startlingly, his judgment also suggests that freedom of testation is not implicated at all, because it does not extend to freedom to testate in a way that is unlawful.Footnote 33 The judgment does suggest that differentiating between public and private testamentary instruments constitutes a distinction without difference, on the basis that both have the same impact on a complainant.Footnote 34 His judgment does not engage in a justification analysis, however, because of the respondents’ determinative concession.Footnote 35 This is unfortunate given the case’s potentially significant precedential effects.
In contrast, Mhlantla J’s separate judgment confronts the conflict between freedom of testation and equality, and explains why equality should prevail,Footnote 36 including because the fideicommissum benefited male descendants solely based on their gender and not because the testator had any personal relationship with the descendants in question.Footnote 37 Notably, Mhlantla J finds that the provision is subject to the Constitution, but also suggests that when courts intervene in private decisions (such as those that manifest in testamentary bequests), “there ought to be a lower level of judicial scrutiny”.Footnote 38 She takes up this line again in Wilkinson, as I discuss below. However, her judgment problematically ties protected grounds to immutable characteristics (taking a narrower approach than prior jurisprudence).Footnote 39 Here, her judgment is unfortunate, not least because some supposedly immutable characteristics (such as gender) can be fluid, and the justification for the law’s protection does not flow from their immutability.Footnote 40 Nevertheless, the overall thrust of her judgment is threefold: first, that systemic patriarchy requires taking a structural approach to addressing disadvantage as it manifests in discriminatory testamentary provisions; second, that such provisions must be tackled through the existing common law public policy test; and third, that while the private sphere is subject to anti-discrimination duties, it should face a lower level of scrutiny.
Victor AJ also takes a systemic approach to substantive equality, emphasizing the (re)distributive aspects of anti-discrimination protections.Footnote 41 She also stresses collective dimensions of equality and ubuntu in “establishing a compassionate society which does not discard the humanity of any of its members”.Footnote 42 However Victor AJ rejects the idea, endorsed by Mhlantla J, that there should be varying levels of scrutiny for the private sphere, seemingly advocating for a more uniform approach to discrimination claims based on the discrimination’s impact.Footnote 43 She recognizes the clash between competing rights but warns against over-emphasizing “individualist, libertarian and neo-liberal definitions of freedom of testation”, or maintaining a rigid public / private divide.Footnote 44 Unlike the other judgments, she specifically draws on section 14 of PEPUDA, which contains the legislative test for determining whether discrimination is unfair, although her analysis of this provision is quite brief. She concludes that the only legitimate purpose served by enforcing the discriminatory clause would be advancing freedom of testation, but concludes that “there can simply be no contest between the raison d’etre (reason for being) of the Constitution, namely the abolition of patriarchy and sexism, and the ‘right’ to freedom of testation”.Footnote 45
Wilkinson v Crawford
Two months after King, the Constitutional Court was again confronted with the tension between freedom of testation and the right to equality. Wilkinson involved the provisions of a private trust deed executed in the early 1950s, which included a clause that if any of the testator’s children and beneficiaries died before the trust’s termination, their shares would devolve to their descendants. At issue was whether descendants include adopted children, given a legislative presumption when the trust deed was executed that adopted children were excluded from inheriting unless explicitly included – and here, the trust deed was silent.
One beneficiary, concerned that her adopted children would not benefit, approached the courts. The High Court found against her, concluding that the testator had intended to exclude adopted children, and that there was no unfair discrimination because the right to equality must be weighed against freedom of testation.Footnote 46 The Court suggested that this balancing should take place in terms of section 36 of the Constitution, the general limitations clause, rather than as part of an unfair discrimination enquiry.Footnote 47 A majority of the Supreme Court of Appeal upheld the High Court’s order, distinguishing between a private family trust deed and a charitable one, and accorded the former more weight to freedom of testation (undergirded by the right to property in section 25 of the Constitution) than to the right to equality.Footnote 48 The Court noted that, first, there remains a sufficiently private sphere that is immune from the prohibition against discrimination; second, no person has any entitlement to benefit from a will; third, public policy should not depend on idiosyncratic value judgments; and finally, applying the right to equality could “lead to a range of insurmountable practical difficulties”.Footnote 49 A minority judgment disagreed, holding that the trust deed’s language did not indicate an intention to exclude adopted children.Footnote 50
The Constitutional Court overturned the Supreme Court of Appeal majority, but was not unanimous.Footnote 51 A majority of the Court (this time, per Mhlantla J) found that the deed could not be interpreted to include adopted children given the legislative presumption, and so must be assessed to determine whether it is discriminatory and out of step with public policy. Her judgment found that excluding adopted children would constitute unfair discrimination (on the ground of birth).Footnote 52 Mhlantla J noted that freedom of testation implicates the rights to property, dignity and privacy, and that public policy must be determined when the trust deed is enforced, rather than when it was executed.Footnote 53 While public policy is readily applicable to public charitable trusts, here the bequest formed part of a private trust without public characteristics, and so “the freedom of testation of the donor ought to be interfered with minimally”.Footnote 54 Reaffirming her approach in King, she found that this does not mean that the provisions of private trusts cannot be tested against constitutional demands; while they are subject to lower scrutiny, they must not be contrary to public policy, in part because no matter how private the trust is, a court (as a public body) will be required to enforce its terms.Footnote 55
Majiedt J dissented, disagreeing that the protected ground of birth extends to adoption and emphasizing the importance of freedom of testation in the private sphere. Flowing from this, Majiedt J’s judgment finds that the exclusion of adopted children from benefiting from the estate constitutes mere differentiation, rather than discrimination. As Bishop and Brickhill compellingly argue, this elision of the context of adopted children is “clearly unsupportable” and at odds with substantive equality and established judicial precedent.Footnote 56 The judgment also draws a sharp distinction between public and private spheres. Testamentary decisions, contends Majiedt J, “are taken in a most intimate, personal sphere – they occur within what this Court has called the person’s ‘inner sanctum’, and within ‘the core most protected realms of privacy’”.Footnote 57 By the Court intervening, it casts judgment on deeply personal choices.Footnote 58 And the distinction between public and private trusts is justified, in Majiedt J’s argument, because public ones are intended to benefit the general public and usually endure over longer periods of time (thereby also prolonging any potential infringement of the dignity of an excluded group of persons).Footnote 59 Conversely, he contends, in the context of private trusts there is no prima facie right to inherit, and the countervailing testator’s rights to privacy, dignity and property insulate the testamentary instrument from excessive court intrusion. Majiedt J’s reasoning echoes the High Court and majority Supreme Court of Appeal judgments that constitutional rights principally protect individuals from undue state influence and coercion, and that it is arbitrary to impose different conditions on a deceased’s estate versus an inter vivos gift.
Jafta J wrote separately in favour of the adopted grandchildren, based on the language in the deed itself, which includes the term “legal descendants”, which he notes is distinctive from the term “biological descendant”.Footnote 60 His judgment delves into the 1937 legislative provision that put in place the presumption that adopted children are excluded, absent a clear intention to the contrary; he found that the conditions for that provision were not met. This conclusion meant that he did not enquire whether the trust deed is discriminatory or interrogate the possible application or development of the common law public policy test. In sum, King and Wilkinson represent a significant and welcome development of South African law on the anti-discrimination duties of private persons. Nevertheless, the judgments leave important questions unanswered.
Normative foundations: Why private testamentary decisions should be constrained by anti-discrimination principles
The judgments, especially King, have already attracted academic commentary, particularly on the balance struck between equality and freedom of testation. Commentary is sharply divided on whether the judgments should be celebrated as a vindication of substantive equality and dignity, or instead criticized as subverting the existing law of succession and sidelining autonomy. In this section, I examine the normative underpinnings for the Court’s extension of anti-discrimination demands into private testamentary decisions by tracking three divergent approaches that emerge from the judgments and academic commentary: (1) an approach that favours insulating the private sphere; (2) an approach that focuses on the impact of the discrimination on the complainant, irrespective of the identity of the discriminator; and (3) an approach that advocates for the private sphere to be regulated, but subject to a lower level of judicial scrutiny.
Insulating the private sphere
The first approach to addressing discriminatory testamentary provisions favours insulating the private sphere, and so also private persons, from legal intrusion.Footnote 61 In this approach, the private testamentary sphere is insulated because personal choice, including choosing who should inherit and on what basis, constitutes an exercise of autonomy. In the tussle between equality and autonomy interests that freedom of testation protects, autonomy should prevail. Rooted in a stricter understanding of the public / private divide, this approach also expresses anxiety about state intervention in intimate spheres.
This reasoning underpinned the King and Wilkinson High Court judgments. In King, Bozalek J rejected the argument that the discrimination was systemic and emphasized the arbitrariness of imposing an anti-discrimination duty to override freedom of testation.Footnote 62 Similarly, in Wilkinson, Dlodlo J asserted that the Constitution primarily guarantees individual freedom from undue state interference.Footnote 63 The judgment drew a parallel between testamentary provisions and inter vivos donations (eg if a person decides to gift property to some but not all of their descendants on discriminatory grounds), arguing that judicial intervention in the former would be arbitrary when it would not occur in the latter. The Supreme Court of Appeal’s majority judgment in Wilkinson echoed this, finding that public policy does not warrant interference in private testamentary instruments and that judicial restraint must be exercised.Footnote 64
At the Constitutional Court level, Majiedt J’s dissent in Wilkinson traverses similar concerns. He emphasizes that testamentary decisions “are manifestations of personal love and affection, loyalties and kinship” and that accordingly, courts should be loath to interfere.Footnote 65 This objection captures a concern that affective relationships (characterized by emotion and intimacy) should not be subject to external directives.Footnote 66 By intervening, Majiedt J argues, a court “may dictate to the testatrix whom she may and may not love”, undermining the substratum of interpersonal relationships.Footnote 67 His judgment also draws a sharp distinction between public and private spheres, arguing for regulation of the former but not the latter to protect an intimately private sphere, where legal regulation of personal choice is impermissible.
While this view did not find favour with the majority, several academic commentators support it. Du Toit, for instance, contends that the public / private divide should be retained (viewing the King judgment as abandoning that divide), on the basis that the state should intervene minimally in private affairs.Footnote 68 He also cautions against using public law norms as “blunt normative instruments” wielded against existing common law principles, including freedom of testation. Sonnekus, favouring an insulated private sphere, goes further, asserting that no unfair discrimination occurs in these testamentary decisions, as no person is prejudiced by another’s benefit and there is no inherent right to inherit.Footnote 69 He characterizes King as unconvincing and questions its precedential effect on testamentary bequests that discriminate on gender for religious reasons. Elsewhere, Sonnekus has asserted that the “apex principle of freedom of testation” should be protected even it excludes beneficiaries on sexual orientation or religious grounds; courts, he argues, cannot interfere “under the pretext of the equality principle”.Footnote 70
However, this stance raises concerns. First, it is unclear why freedom of testation should be an apex principle or why an almost unfettered approach should win out. This is evident in the seemingly question-begging example of the religiously motivated testator. True, a court would have to add freedom of religion to the mix of factors to balance, but Sonnekus treats the outcome as predetermined – ie that the testamentary condition would be upheld and enforced – rather than as instantiating a genuine clash. Second, the expressive harm of discriminatory testamentary provisions is another important, albeit incomplete, consideration. Ahmed argues that a will constitutes an “unrivalled legal instrument of self-expression” that communicates messages of belonging or exclusion.Footnote 71 In King and Wilkinson, the testamentary provisions demeaned women (in King) and adopted children (in Wilkinson), expressing that they are worth less and do not belong.Footnote 72 However, these expressive harms also arise in inter vivos donations that are generally not legally regulated. Ahmed’s response, which emphasizes the especially communicative aspect of will-making, thus seems partial.Footnote 73 While expressive – or, as Fraser would characterize them, recognition – harms seem to justify the judgments, there is perhaps also a meaningful role played by courts’ reluctance to enforce discriminatory provisions (which they would not generally do for discriminatory inter vivos donations).Footnote 74
Third, the stance favouring an insulated private sphere relies on a contested understanding of autonomy. Conversely, Harding argues that in King there is a negligible impact on freedom of testation, because discriminatory disinheritances are not valuable exercises of autonomy and thus ought not to be protected. Put differently, Harding suggests that even if autonomy is valued, not all exercises of it warrant legal protection.Footnote 75 Similarly, Osman and Effendi note that King aligns with, rather than departs from, existing jurisprudence: freedom of testation has always been subject to limitations, and adjudication often turns on specific facts.Footnote 76 Moreover, this stance seemingly equates autonomy with negative liberty – ie constraining state intervention – and presumes an atomistic individual.Footnote 77 Instead, as Nedelsky argues, a relational account of autonomy recognizes that individuals are constituted by relationships, which can foster rather than impede autonomy.Footnote 78 This suggests that legal regulation can structure relationships to advance autonomy by dismantling hierarchies that drive systemic inequalities. This is all further complicated when we consider critiques of autonomy-based accounts, which suggest that what appear to be independent choices may in fact be determined – or at least shaped – by extant power structures.Footnote 79 As Bhatia notes, testamentary decisions such as those in King and Wilkinson constitute “classic example[s] of an individual act having specific salience because of the institutional, or systemic, relational context within which it is embedded”.Footnote 80 A relational egalitarian approach, which views social relations as a site of justice, suggests that equality claims exert force not only on the state but also on society.Footnote 81
Impact as determinative of duty?
If the first view would insulate the private sphere from legal regulation and preserve the public / private divide, at least some of the Constitutional Court judges in King and Wilkinson seemingly took the opposite view, emphasizing the divide’s permeability. This second approach treats private actors similarly to the state, recognizing that discrimination’s impact on a complainant can be identical, regardless of its source. Aspects of this line of reasoning emerge from Jafta J’s majority and Victor AJ’s separate judgment in King. Jafta J stresses that whether a testamentary instrument is public or private, if it has the same impact on the complainant, it is equally constitutionally invalid.Footnote 82 Anything to the contrary, he suggests, implies that the Constitution does not reach the private sphere. Victor AJ’s judgment does not render the public / private divide completely irrelevant, but rejects applying differing levels of scrutiny to discriminatory conduct in these spheres.Footnote 83 Instead, highlighting the question of power, she notes that discrimination has the same consequence – or impact – on a complainant whether it originates in public or private spheres, necessitating constitutional scrutiny.Footnote 84
However, we should not be too quick to characterize either of these judgments as crudely endorsing a view that private duties should wholly mirror state duties. In King, both Jafta J and Victor AJ focus on the discrimination’s impact, but impact is distinct from justification, and both matter in determining fairness. Jafta J’s judgment did not reach the justification leg at all, because the parties conceded that the discrimination was unfair. Arguably, his judgments in King and Wilkinson favour a casuistic approach, avoiding an overarching test and focusing on case specifics. His King majority found that the respondents’ concession disposed of the dispute, and so he declined to engage in a fairness analysis.Footnote 85 In his separate judgment in Wilkinson, he contended that the dispute turns on a close reading of the deed’s language, avoiding broader debates on balancing freedom of testation and equality.
While Victor AJ emphasizes that discrimination’s impact may be identical irrespective of the discriminator, she also acknowledges the clash of interests between private persons. She suggests that PEPUDA’s section 14 should accommodate this and engages with that enquiry, albeit briefly. She notes that section 14 is “a hybrid test which incorporates elements of the fairness enquiry from Harksen whilst also incorporating elements of proportionality that resemble a limitation analysis”.Footnote 86 Hers is the only one of the six cumulative Constitutional Court judgments to engage with section 14. She notes that women are a vulnerable group who have been systemically discriminated against, perpetuating patterns of disadvantage, and that the countervailing legitimate purpose does not outweigh this.Footnote 87 Unlike Jafta J, she does not deny that freedom of testation is implicated; however, she concludes that it is outweighed by the importance of equality in this case.Footnote 88
In any event, it would be a mistake to straightforwardly mirror state and private anti-discrimination duties. Instead, certain cases of discrimination that would be clearly prohibited if the state were the respondent may be legally permitted for private persons, as the basis for justifying discrimination is typically larger. That determination takes place not as a threshold enquiry as to whether the private person is a duty-bearer in the first place but instead as a downstream enquiry as to whether the discrimination is nevertheless justified and fair.
Intensity of review: Variations on a theme
Finally, I consider a third approach that varies the intensity of review based on the discriminator’s identity. Mhlantla J takes an intermediate view between the two positions earlier (insulating the private sphere or taking impact to be determinative) in her separate King judgment and the Wilkinson majority judgment. In both, she ultimately falls on the side of equality, finding that private testamentary provisions are subject to anti-discrimination controls. However, she finds these controls should operate through the common law public policy test (rather than the Constitution or PEPUDA). In King, where she did not attract a majority, she recognized that private testamentary bequests fall within intimate personal relationships, warranting a “lower level of judicial scrutiny”, given the impact on the right to privacy.Footnote 89 This does not render the private sphere immune from scrutiny; instead, given the clear discrimination on the ground of gender, the bequest was plainly contrary to public policy. In Wilkinson, her majority judgment asserts that courts should interfere minimally with freedom of testation for private trusts and that “stricter scrutiny” applies to public charitable trusts.Footnote 90 She leaves open how these standards differ. However, as her judgment in Wilkinson was the majority (and Jafta J’s majority in King principally turned on the respondents’ concession, and so its ratio decidendi is not ultimately inconsistent with her finding), it establishes differing legal standards for public versus private discriminatory acts and suggests broader scope for justification by private persons.
This is a novel development, and at first blush, applying a lower intensity of review to private duty-bearers is quite attractive. There is an intuitive case for subjecting private conduct, especially individuals’ personal decisions, to attenuated scrutiny, as compared to discriminatory conduct by the state, and varying the intensity of review depending on context is not an alien concept in South African law. There is a well-established seam of doctrine in South African administrative law that varies the intensity of judicial review applied to exercises of public power, dependent on context and the appropriate level of deference.Footnote 91
However, unlike US anti-discrimination law, where scrutiny levels vary by protected grounds, South Africa’s equality jurisprudence has not yet seriously engaged with this approach.Footnote 92 Indeed, in an early equality case, Prinsloo v Van der Linde, the Court cautioned that as “a simplistic translation from other countries into our equality jurisprudence of formulae, modes of classification of degrees of scrutiny might create more problems than it solves”.Footnote 93 McConnachie argues for a baseline intensity of review when assessing state justifications for discrimination.Footnote 94 However, his analysis is framed around the state as duty-bearer, rooted in a culture of justification that applies more obviously to the state than to private persons. Conversely, Mhlantla J’s suggestion is that the intensity of review is dependent on the duty-bearer (rather than on the nature of the power that is exercised, as in administrative law). Further, the varying intensity of a review can capture the differences between variegated kinds of private actors – so discriminatory conduct by a corporate entity pursuing profit can be subject to more searching scrutiny than the personal decisions taken by individuals in more intimate spheres.
But varying the intensity of review does not resolve the difficult questions courts face when adjudicating private anti-discrimination duties. All it does is set the resolution back a step: a court will still need to scrutinize the justification raised to defend discriminatory conduct, and varying the intensity of this scrutiny does not exempt private actors from regulation. Ultimately, a balancing exercise remains necessary. Given this, it is not entirely clear what extra work the standard of a lower intensity of review does, aside from, perhaps, reframing the weight a court may accord during the justification process to a private person’s competing interests.
Resurrecting the form of the application debate
In the previous part, I examined the normative commitments animating the Court’s judgments. I suggested that substantive equality (with relational dimensions) and a non-atomized understanding of autonomy play important roles in shaping our comprehension. I next consider how these considerations should be applied, revisiting contested adjudicative questions of subsidiarity and their direct or indirect application.
The King and Wilkinson judgments reveal significant divergences regarding how the anti-discrimination duty should be given effect. Mhlantla J developed the common law (through the public policy test) to extend it to private testamentary instruments; she held that the appropriate approach is not to insulate the private sphere but to have a variable intensity of review, with public policy doing the heavy lifting through indirect application of equality demands. Her approach, then, is that there is a choice between either testing the provisions under the common law or applying PEPUDA, and that opting for legislation would neglect public policy and fail to develop the common law.Footnote 95 Jafta J also treats the choice as binary, explaining in King that the public policy claim is discrete from the equality claim, in part because the former engages the right to equality, while the latter engages with the value of it.Footnote 96 However, this reasoning sidesteps the issue – particularly given his finding that freedom of testation was not engaged at all. On the form of application, Jafta J’s King majority is curious: he finds that section 9 and PEPUDA should apply and that the common law need not be developed as freedom of testation has always been circumscribed by what is lawful. Victor AJ’s judgment in King engages with the doctrine of subsidiarity; she concludes that eschewing PEPUDA would subvert the doctrine of subsidiarity, although elsewhere she recognizes that substantive equality should form part of the public policy test.Footnote 97 Ultimately, she finds that direct horizontal application is to be used given the nature, intensity and history of the right to equality and the dual dangers of insulating the private sphere and undermining substantive equality that, by implication, indirect application via the common law would risk.Footnote 98 The concern with this approach, however, is that overlooking and failing to give effect to the common law gives rise to exactly the risk that she seeks to avoid: that some areas of law are insulated from constitutional precepts.
South Africa’s constitutional scheme envisages that private persons do owe each other constitutional obligations in appropriate circumstances. Typically, this should be through indirect application as mediated through more subsidiary, discrete legal rules, rather than through direct reliance on the right itself, for these rules can better account for the competing interests which may be at play in private legal relationships.Footnote 99 This approach avoids perpetuating parallelism, ie the idea that the common law exists in parallel to the Constitution and its demands, and instead aims for constitutional diffusion across all areas of the law.
However, we see two pressure points in the judgments, exerted from competing directions: first, the idea that the common law has a vanishing role here in the presence of PEPUDA and section 9 of the Constitution (Jafta J and Victor AJ), and conversely, that the common law ought to be developed as the natural home for this enquiry, sidestepping the specific legislation enacted to give effect to the right to equality (Mhlantla J). In truth, this framing rests on a false dichotomy. The choice need not be between entirely ignoring the common law, despite the existence of a well-established rule that regulates unlawful testamentary bequests, or subverting the legislative scheme – with the associated separation of powers and institutional risks – by failing to use PEPUDA. Instead, a harmonized approach should be taken.
Again, academic commentators have debated these aspects of the judgments. Bishop and Brickhill prefer Mhlantla J’s majority in Wilkinson for its clear protection of adopted children’s rights and its weaving together of the principles of equality, freedom of testation and public policy.Footnote 100 Van Zyl criticizes the King majority for diminishing “the pertinent role of public policy in our constitutional democracy”, arguing that the right to equality is not core in certain areas of private law.Footnote 101 She contends that public policy should be informed by the norms of the community, and sounds some scepticism about whether the community would find that equality trumps values of dignity and freedom (including freedom of testation).
Basson advocates for a transformative approach to adjudicative subsidiarity.Footnote 102 He argues against pitting equality, freedom and dignity against each other, in favour of a non-atomized approach to autonomy and a structural understanding of discrimination. He cautions against a mechanistic balancing exercise, and argues for eschewing the common law in favour of PEPUDA, due to the latter’s greater protection for disadvantaged individuals and groups and because of institutional legitimacy. Basson rejects the idea that the common law, which embodies doctrinal principles such as fairness and public policy, is coextensive with PEPUDA, on the basis that those doctrinal principles have largely been developed to embody classically liberal (and thus also individualistic) ideas, which are at odds with a commitment to transformative substantive equality. On King, Basson notes that the judgments each take varying approaches to subsidiarity, arguing that all three judgments used “unstable” manoeuvres and fail to provide clarity. Instead, he favours direct application and is wary of Mhlantla J’s use of the common law’s public policy test, arguing that it is a judicial subversion of legislative norms. He is also suspicious of attempts to elevate freedom of testation to a “fully fledged right”, particularly where it must be constrained by the expressly constitutional protected anti-discrimination right. Jafta J’s judgment, he notes, circumvented dealing with the difficult issue by finding that freedom of testation does not permit unfair discrimination. As Basson rightly points out, section 14 of PEPUDA requires robust judicial reasoning in each case, rather than blanket assertions of unfairness. Basson favours Victor AJ’s approach, praising her use of the doctrine of subsidiarity to vindicate transformative substantive equality while avoiding reinscribing the public / private divide. However, he cautions that her judgment displays too much optimism in PEPUDA’s ability to resolve unfair discrimination disputes easily, warning instead that where more robust defences of discrimination are raised, courts will have to delve into the section 14 unfairness analysis.
The recent academic commentary on the judgments takes place against the backdrop of long-running and labyrinthine disputes about the form of horizontal application, and the contested and continuing role of the common law in a constitutional democracy.Footnote 103 Recently, Boonzaier has dissected a worrying trend of the Constitutional Court jettisoning an indirect or mediated approach (where the court avoids a bifurcation of the constitutional system and the common law, instead developing the common law to bring it into alignment with the Constitution) in favour of direct application that displaces the existing common law rules and remedies. This, he argues, is unfortunate: constitutional and common law perspectives “are meant to be cooperative and mutually illuminating”.Footnote 104 In the context of contract law, Bhana too has argued for a harmonized approach, folding the normative provisions of PEPUDA into a contractual enquiry.Footnote 105 These insights apply with equal force to the law of succession. In Bhana’s account, the common law can and should be developed in line with PEPUDA. She argues persuasively for the importance of reconciling the common law with PEPUDA, rather than the blunt application of the latter to contracts in a way that scrubs out existing common law principles. This, she notes, avoids creating parallel systems of law.
In my view, a harmonized approach, which employs PEPUDA to give further content to public policy, is principled and correct. A harmonized approach does not pit the two areas of law against each other and avoids proliferating parallel systems of law, bringing the common law in line with constitutional demands without jettisoning it entirely. This is consistent with the Constitutional Court’s own precedent that the common law and statute must be read in harmony.Footnote 106 PEPUDA is not necessarily intended to comprehensively cover the field, but it is the legislation put in place to regulate unfair discrimination disputes, including the regulation of testamentary bequests. However, this need not necessarily mean that it operates to the exclusion of public policy in determining a dispute, to entirely displace the common law rather than working in harmony with it.
However, how a harmonized approach works – and specifically, how the legislative unfair discrimination enquiry interacts with the existing common law cause of action – is complicated.Footnote 107 One possibility is for PEPUDA’s section 14 to inform public policy, harmonizing the analysis demanded by PEPUDA within the existing common law frame, rather than bifurcating this enquiry, with the concomitant perils of parallelism. This still introduces more substantive content to the public policy enquiry within the law of succession enquiry, without undermining the legislative scheme. A harmonized approach in this context requires a greater role to be accorded to section 14 of PEPUDA in shaping the public policy test within the common law of succession.
South African law puts in place a more extensive regime for private people’s anti-discrimination duties compared to other jurisdictions. In principle, all persons as well as the state are duty-bearers, eliminating any presumption that private persons or the private sphere are insulated from the equality obligation. However, while the anti-discrimination duty applies to both the state and private persons, the justification for discrimination may be broader for private persons.Footnote 108 Put differently, the creation of a wide class of duty-bearers extends the reach of anti-discrimination law, but the availability of justifications or defences potentially attenuates its application. However, a litigant cannot simply appeal to courts to be less intrusive in more intimate spheres – many egregious human rights violations take place within deeply personal spheres. Instead, a proper analysis of the competing rights and interests must be undertaken, through the section 14 justification stage.
The enquiry into fairness, then, is the tool “for sifting permissible from impermissible discrimination”.Footnote 109 Section 14 of PEPUDA governs this determination, providing several factors to inform the assessment; it was partially based on the judicial factors proposed in Harksen v Lane, as well as components of the constitutional limitations analysis.Footnote 110 Although insufficiently tested in doctrine, section 14 is wide enough to “accommodate a consideration of other rights”.Footnote 111 It articulates various factors, including context, the impact of the discrimination on the complainant and extant patterns of disadvantage, and whether the discrimination has (and achieves) a legitimate purpose. The role of section 14 also needs to be developed against section 13 of PEPUDA, which puts in place a presumption that discrimination on protected grounds is unfair and places the burden of proof on the respondent once the complainant has established a prima facie case.
Sadly, section 14 has received very little attention in judgments to date.Footnote 112 Litigants have tended not to rely on it to justify discrimination even where there are apparently competing rights at stake. This is compounded by confusion about whether claims of discrimination based on race, gender or disability should proceed under the general prohibition in section 6 of PEPUDA or the more specific provisions in sections 7, 8 and 9. Where South African courts have referred to section 14, it is often in passing or in state-centric contexts.Footnote 113 Consequently, courts have yet to develop a clear test for balancing competing rights, resulting in casuistic and fact-dependent enquiries. While context sensitivity is critical, some more principled guidance is needed.
Basson’s recent attention to section 14 of PEPUDA proposes guidelines for understanding the fairness enquiry in the context of recognizing poverty as a ground of discrimination.Footnote 114 He argues that section 14 does not require mechanistic balancing or an “anything goes” value judgment; instead, he separates out considerations of the impact of the discrimination from its possible justification, arguing that transformative substantive equality requires that the courts consider interrelated factors in a proportionality analysis which protects the rights of disadvantaged persons.
Turning again to the provision itself, it is true that section 14 incorporates a proportionality assessment into the determination of unfair discrimination. It includes factors such as whether the discrimination has a legitimate purpose, whether and to what extent it achieves its purpose, and whether there are less restrictive and less disadvantageous means to achieve the purpose.Footnote 115 Section 14(1) provides that discrimination is not unfair if it involves measures to protect or advance disadvantaged persons. Section 14(2) and (3) outline the test for determining fairness, including context and other relevant factors.Footnote 116 These can be broadly divided into two categories: those focused on impact (sub-sections (a) to (e)) and those concerned with justification (sub-sections (f) to (i)).Footnote 117 Reading section 14 as a whole, the overall considerations can be grouped into varying categories: first, considerations that go to redressing inequality and so rebut unfairness; second, complainant-oriented considerations (which focus on impact); and third, defendant-oriented considerations (which focus on justification and proportionality). The impact factors examine the discrimination’s effect on the complainant or protected group, while the justification factors consider the discriminator’s reasons, purpose and behaviour, incorporating some wording from section 36 of the Constitution.
Section 14 of PEPUDA exemplifies a legislative balancing enquiry that requires the courts to provide further adjudicative and substantive content. The legislature has set an overarching framework, mandating courts to determine the appropriate balance in specific cases in a structured yet non-mechanical manner. For this reason, the clash between rights need not collapse into a balancing enquiry that straightforwardly sets unencumbered freedom (that is, negative liberty interests as understood, shorn of any institutional or structural context) in opposition to equality.Footnote 118 This need not mean, however, that a balancing enquiry (which incorporates privacy and autonomy interests) is not at all germane. Liebenberg and Kolabhai put it aptly: “Considerations of private autonomy would enter the picture, not in terms of the application analysis, but rather in terms of the substantive balancing between various rights.”Footnote 119 Such balancing can accommodate structural and institutional concerns, considering both the relational context of the parties and the underlying demands of substantive equality without presuming parity of power among individuals. This structured balancing enquiry should form part of a harmonized approach that gives content to section 14 of PEPUDA and develops the common law public policy test, providing a more nuanced and contextually sensitive framework for addressing discrimination in both public and private spheres.
Applying the harmonized approach: King and Wilkinson revisited
To concretize the argument for a harmonized approach, this section applies section 14 of PEPUDA within the common law public policy test to the facts and reasoning in King and Wilkinson.Footnote 120 This ensures that the unfair discrimination enquiry is housed within the established common law framework while giving full effect to PEPUDA’s provisions and the constitutional imperative to develop the common law in line with sections 8(3) and 39(2) of the Constitution.Footnote 121
The overarching question in cases challenging testamentary bequests is whether the bequest is contrary to public policy.Footnote 122 There are multiple grounds on which a bequest may be deemed unenforceable under the public policy test, but one clear impetus for unenforceability would be that the bequest constitutes unfair discrimination. This then triggers the PEPUDA enquiry, which first examines whether discrimination is established and, if so, places the burden on the discriminator to prove that the discrimination is fair in terms of section 14. Within this enquiry, principles such as negative autonomy and freedom of testation remain important, but are subject to structured scrutiny. The weight of these principles must be calibrated in light of competing constitutional commitments, most obviously substantive equality. In this balancing enquiry, then, autonomy and testamentary freedom are not disregarded, but must yield where their exercise entrenches systemic disadvantage, ensuring that private law doctrines align with constitutionalism. Accordingly, if the discrimination is ultimately found to be unfair, the bequest is rendered contra boni mores and cannot be enforced. This harmonized approach thus provides courts with a clearer framework for balancing these competing interests, while remaining sufficiently supple to accommodate the contextual circumstances of each case. It also embodies adjudicative subsidiarity by drawing on PEPUDA rather than directly invoking section 9 of the Constitution, while simultaneously ensuring that the common law is not sidelined but developed in accordance with constitutional rights and values.
This approach is indebted to Bhana’s work on the intersection of contract and anti-discrimination law.Footnote 123 Bhana argues that equality law should not be treated as a parallel regime but instead must be situated within the common law framework that governs the relevant area. In the contractual context, she contends that courts should assess an allegedly discriminatory contractual clause within the public policy test, with PEPUDA’s unfair discrimination enquiry providing the structured framework for determining enforceability. This prevents the common law from becoming obsolete while ensuring that anti-discrimination principles are meaningfully integrated into private law adjudication. A similar logic applies in the law of succession: the anti-discrimination duty constraining testamentary freedom must be housed within the existing common law public policy test, rather than treated as a separate and independent enquiry.
In King, the fideicommissum that excluded female descendants from inheriting was prima facie discriminatory on the grounds of gender. Under the harmonized approach, the court would determine whether the clause is contrary to public policy by engaging in the structured PEPUDA enquiry. The complainants would need to establish that the clause differentiated on a prohibited ground (gender), in turn drawing on section 8 of PEPUDA, which gives specific content to the prohibition of unfair discrimination on the ground of gender.Footnote 124 This is clearly met here, given the express exclusion and South Africa’s patriarchal history of ownership. The burden would then shift to the opposing party to justify the discrimination under section 14, which considers factors such as whether the clause perpetuates patterns of disadvantage. Given the historical exclusion of women from inheritance, the justification would likely fail, notwithstanding freedom of testation, leading to a finding that the bequest was unfairly discriminatory. Consequently, under the public policy test, it would be deemed contra boni mores and unenforceable. This approach not only aligns with the Constitutional Court’s jurisprudence but also clarifies the legal reasoning by structuring it within the common law.
In Wilkinson, the exclusion of adopted children in the trust deed similarly required assessment under the public policy test. The first question was whether the bequest was contrary to public policy; given that exclusion based on birth status can constitute unfair discrimination, the court would then turn to the PEPUDA enquiry. The complainant would need to show either that adoption falls under the listed ground of birth or is an analogous ground; as I noted earlier, the former is the more potent argument, and Majiedt J’s judgment in Wilkinson was especially flawed in its treatment of this aspect of the unfair discrimination enquiry. The burden would then shift to the discriminator to justify the exclusion under section 14. Excluding adopted children reinforces historic legal distinctions that treat adoptive relationships as inferior to biological ones, with both material and dignitarian consequences which run against jurisprudence that explicitly affirms the diversity and equal worth of all family forms.Footnote 125 The discrimination would then be found unfair, rendering the provision contrary to public policy and unenforceable.
One possible objection to this harmonized approach is whether it leads to unnecessary over-determination of the outcome. If the PEPUDA enquiry is nested within the public policy test, does this mean that either enquiry alone would have sufficed? And does it make any difference, given that the analysis sketched above ultimately results in the same outcomes as the Court’s, in both judgments? The answer is no. Rather than proliferating parallel legal routes, this approach ensures that both legal frameworks retain their role and relevance. Treating them as distinct risks bifurcating equality jurisprudence from private law, leading to ad hoc adjudication that fails to develop a coherent body of law. Housing the anti-discrimination enquiry within the common law test prevents this fragmentation while preserving the force of both legal frameworks.
This argument is reinforced by the distinction between the orders in King and Wilkinson. In King, the clause was declared inconsistent with the Constitution and PEPUDA, with no explicit reference to the common law.Footnote 126 In Wilkinson, however, the court found that the clause was contrary to public policy and unenforceable, without invoking PEPUDA.Footnote 127 This discrepancy illustrates the problem of parallelism: courts have at times relied solely on constitutional and statutory principles and at other times exclusively on the common law, without integrating the two. The harmonized approach resolves this by ensuring that PEPUDA informs the public policy enquiry while maintaining the common law as the doctrinal home for adjudicating testamentary discrimination disputes.
Ultimately, this approach aligns with the Constitutional Court’s recognition that there is only one system of law in South Africa.Footnote 128 PEPUDA does the substantive work of structuring the unfair discrimination enquiry, but the common law serves as the adjudicative framework that houses both the enquiry and the remedy. By embedding the anti-discrimination duty within the public policy test, this harmonized approach provides conceptual clarity, doctrinal coherence and a principled basis for adjudicating testamentary discrimination claims.
Conclusion
This article has evaluated two judgments from South Africa’s Constitutional Court which affirm that anti-discrimination duties constrain private testamentary decisions. King and Wilkinson represent a significant development of South African law on private anti-discrimination duties. The multiple judgments of the Court evince divergent approaches to the regulation of private discrimination; however, different considerations may be apposite when duties are imposed on private persons rather than the state, or in the private rather than the public sphere. These judgments break decisive ground in South African succession law, reinforcing that private testamentary decisions are not immune from constitutional scrutiny. Reading the two judgments together, the Court has endorsed that the private sphere cannot be insulated from the right to equality, but also that it is nevertheless subject to a lower scrutiny of review.
I examined the judgments through dual lenses: first, a normative lens that assesses why the Court concluded that private testators cannot unfairly discriminate in private bequests, and second, an adjudicative lens that takes up the question of how this anti-discrimination duty should be imposed. The persistence of parallel legal paths – where courts shift between applying the public policy test, section 14 of PEPUDA or constitutional principles – creates unnecessary doctrinal fragmentation. I began by charting how freedom of testation has long been circumscribed by the boni mores test, which requires that testamentary decisions be constrained by public policy. I then considered the implications of substantive equality and South Africa’s expansive approach to the anti-discrimination duty of private persons on freedom of testation and the public / private divide. South African anti-discrimination law stems from a history characterized by the oppressive, dominating and discriminatory regulation of all spheres, including not only the public but also the private (and sometimes, as in the regulation of sex and marriage, the deeply personal). This created and perpetuated structural inequality that cut across all sectors of society, creating pervasive systems of discrimination that could not be dislodged by the regulation of public spheres or decisions alone. Normatively, substantive equality rejects formal equality’s implicit presumptions that individuals have parity of power, and that the status quo is neutral with respect to the allocation of advantages, rights and basic goods. Preserving a rigid public / private divide – including by insulating testators from anti-discrimination duties in their private bequests – risks calcifying an unjust status quo and eliding inequalities and abuses of power by non-state actors. Put differently, substantive equality, including commitments to relational egalitarianism and to a thicker conception of autonomy that does not equate to negative freedom alone, provides the basis for the broad scope of the anti-discrimination duty. Consequently, anti-discrimination law reaches into what is classically thought of as the private sphere.
However, this does not mean that there is no conceptually useful role played by concepts of autonomy and privacy, or that there may be difficult cases that require careful and principled balancing.Footnote 129 Substantive equality, then, mandates an approach to the public / private divide that does not differentiate between a legally regulated public sphere and a legally insulated private one, but instead allows for autonomy and privacy interests to play a role when balancing competing rights, as part of the justification leg of the unfair discrimination enquiry.
I then evaluated the judgments by considering the adjudicative question of how, doctrinally, these legal changes should be effected. This part of the article surveyed the form of the application of horizontal anti-discrimination duties, which has been and remains a vexed site of jurisprudence and academic commentary. I argued against the proliferation of parallelism, which would have the common law existing in parallel to the Constitution and the rights-enacting legislation (such as PEPUDA) that it requires. Instead, I advocated for a harmonized approach that gives force and doctrinal life to both section 14 of PEPUDA, which sets up the unfair discrimination enquiry and has to date been neglected in doctrine and commentary, and the extant common law enquiry. By adopting a harmonized approach to adjudicating King and Wilkinson, this article illustrates how the structured framework of section 14 of PEPUDA, applied in concert with the public policy test, can clarify and systematize judicial reasoning, preventing a parallelism that risks reinforcing the hierarchies that the Constitution seeks to dismantle.
In sum, I argued that the King and Wilkinson judgments are welcome, reinforcing that the private sphere is not insulated from scrutiny, but that they nevertheless leave some important questions unanswered. What exactly this level of scrutiny demands, and the harmonized role of both section 14 of PEPUDA and the common law public policy test, remain under-determined, however, and must be developed in subsequent jurisprudence. While these judgments establish crucial precedent, the challenge moving forward is to ensure consistency in adjudicating private anti-discrimination claims, avoiding doctrinal fragmentation and reinforcing a principled and structural application of equality law.
Competing interests
None
Open access
